List your company

Multi-turn Calibration on TriviaQA

5.04ECE@1MTCal
3.498413.904224.3134.7158Apr 7, 2026

Evaluation results

MethodLinks
MTCalBackbone=Llama3.1-8B-Instruct2026.045.042.313.2920.012.94
SLBackbone=Llama3.1-8B-Instruct2026.046.4326.419.3622.179.27
DCalBackbone=Llama3.1-8B-Instruct2026.046.621.298.3622.558.05
MTCalBackbone=Qwen2.5-7B-Instruct2026.047.723.054.8321.214.97
DCalBackbone=Qwen2.5-7B-Instruct2026.047.9814.489.9122.748.71
MTCalBackbone=Gemma2-9B-it2026.048.339.154.6913.836.43
PSBackbone=Gemma2-9B-it2026.049.319.9911.3216.6110.39
SLBackbone=Gemma2-9B-it2026.0410.215.467.7915.877.22
PSBackbone=Llama3.1-8B-Instruct2026.0412.9326.4411.4623.167.27
DCalBackbone=Gemma2-9B-it2026.0414.6716.4215.4116.0415.53
SCBackbone=Qwen2.5-7B-Instruct2026.0415.0425.5118.8222.912.94
PSBackbone=Qwen2.5-7B-Instruct2026.0417.2228.8918.6325.715.42
SCBackbone=Llama3.1-8B-Instruct2026.0419.7728.125.5228.8820.67
VerbalBackbone=Llama3.1-8B-Instruct2026.0421.9641.7225.8927.7922.94
P(True)Backbone=Llama3.1-8B-Instruct2026.0422.8732.4723.2925.521.55
SLBackbone=Qwen2.5-7B-Instruct2026.0424.133.3822.2725.5421.72
P(True)Backbone=Gemma2-9B-it2026.0425.7227.417.5817.8116.84
VerbalBackbone=Gemma2-9B-it2026.0426.4130.9228.826.1223.05
SCBackbone=Gemma2-9B-it2026.0430.7136.2246.7940.5236.4
P(True)Backbone=Qwen2.5-7B-Instruct2026.0431.3838.2727.3428.6126.17
VerbalBackbone=Qwen2.5-7B-Instruct2026.0443.5858.2339.4439.8131.1